A recent Wall Street Journal analysis, “Why Big Tech’s AI Spending Is $3 Trillion Higher Than It Seems”, estimates that the AI investment boom includes roughly $3 trillion in spending and contractual commitments beyond what's currently visible on corporate balance sheets. The centerpiece of the WSJ article is an iceberg visualization: AI spending already recognized on company balance sheets appears above the waterline, while trillions of dollars in leases and purchase commitments not yet recorded sit below it. Although the $3 trillion estimate covers nine companies, the visualization focuses primarily on Amazon, Meta, Microsoft and Alphabet.

The headline number is striking, but “off-balance-sheet” should not be confused with improper accounting or disclosure. Under Generally Accepted Accounting Principles (GAAP), purchase commitments often do not immediately create a liability. For instance, a company may not need to recognize a liability until a lease commences or equipment is delivered. Until then, the commitment may appear as a footnote or MD&A disclosure rather than a balance-sheet obligation. Nor are these obligations necessarily hidden. The WSJ could assemble its estimate precisely because the companies reported these arrangements in their SEC filings.

Still, concerns about the scale and structure of AI spending are justified. The industry increasingly depends on a small group of AI labs buying vast amounts of computing capacity from the same companies that invest in them. Some arrangements involve related parties, while others have been described as circular financing: a company invests in an AI developer, which then uses part of that funding to purchase the investor’s products or services. Bloomberg has warned that these links could magnify losses across the industry if AI demand disappoints. The IMF has similarly warned that circular financing increases interconnectedness and could amplify adverse shocks.

The underlying concern is also whether current valuations can be sustained if AI labs fail to generate enough cash to cover the enormous amount of computing capacity they have agreed to buy. A sharp decline in AI-company valuations could affect hyperscalers, chipmakers, data-center developers, lenders, and investors at the same time.

These concerns have revived comparisons with Enron, particularly in commentary about circular financing and off-balance-sheet obligations. The comparison is tempting but misses an important distinction. Enron used off-balance-sheet entities and other accounting practices to conceal debt and losses and mislead investors about its financial condition. The AI commitments the WSJ examined are disclosed in SEC filings, and many remain off the balance sheet because GAAP does not yet require recognizing them as liabilities.

AI-related disclosure can certainly be improved, particularly around commitments and guarantees. If the AI boom ends badly, these obligations could make the losses larger. But the problem would not necessarily be that companies improperly kept obligations off their balance sheets or disclosed leases in footnotes rather than recognizing them immediately as a balance-sheet liability. In my view, loosely defined AI monetization metrics are more likely to inflate valuations than prominently disclosed leases and purchase commitments that may never fully materialize.

The WSJ analysis is therefore a useful measure of scale, but evaluating the $3 trillion figure requires considering:

• Comparability: Companies may define “purchase commitments” differently or include different types of contracts.
• Cancelability: Contractual rights may reduce or eliminate obligations before payment, affecting how binding the headline amounts are.
• Timing: Some commitments are due within one or two years, while others extend for decades.
• Time value of money: Long-dated payments may need to be discounted to present value to be meaningfully compared with obligations already measured on that basis.

This is an abridged version of the analysis. The full post — available to Deep Quarry subscribers — examines how those commitments are defined and how binding they are.

Investment, Tax and Legal Disclaimer: This article is for informational purposes only and does not constitute investment, tax or legal advice. The content contained herein is not to be relied upon as the basis for any investment or other decision. Nothing herein should be construed as a solicitation, recommendation, endorsement, or offer to buy or sell any particular security, product, or service. The author has not taken into account the specific investment objectives, financial situation, or particular needs of any specific person who may read this material. Investing involves inherent risks, and there can be no guarantee that any investment or company mentioned will be suitable or profitable for any investor's investment portfolio. Readers are strongly advised to conduct their own thorough research and consult with a qualified and licensed financial professional and legal counsel before making any investment decisions. Past performance is not indicative of future results. 

Opinion Disclaimer: The opinions and views expressed in this article are those of the author and the parties quoted and not necessarily those of The National Law Review or its Guest Contributors.